Policy implications

It is almost 10 years since the original policy review of ambulance service performance standards and the recommendation that alternative methods of

management for low priority or non urgent category C calls be developed1. For most of that time category C calls have remained within the response time performance framework although the need to explore other options of response, and in particular telephone advice and better integration of the ambulance service and NHS direct call handling services has remained a key policy objective5,29. It has been estimated that up to 40% of 999 calls do not require an emergency ambulance and that alternative responses could both save ambulance resources and provide a better service for patients. It was only in 2004 that category C calls were removed from response time performance targets and ambulance and commissioning services given responsibility for developing local standards and response options for these calls24. The provision of clinically appropriate response options for category C calls is seen as a key strategic objective for modernising ambulance services as set out in the policy document “Taking Healthcare to the patient”6. Enhanced call handling, including providing telephone advice and support to help patients make the right choice for their needs, is highlighted as one of the main developments required to achieve these objectives. This policy initiative also recognises that to for such a service to function effectively patients should only have to make one telephone call and that referral pathways need to be in place to direct patients to appropriate care. Both of these principles are confirmed by the findings of this study and, importantly, we have shown that the consequences of not setting these processes in place results in a service of limited value to a very small number of patients.

The use of telephone advice and assessment as one alternative method of management for category C calls has been explored for a number of years and preliminary studies found the service to be safe and acceptable13. Early estimates suggested that up to 12% of 999 calls could be managed in this way with a

comparable reduction in ambulance service responses. This study has shown that in practice this level of reduction in ambulance responses is unrealistic. Study site 1 did show a potential referral rate of 13% of 999 calls. However the combination of

restricted operational hours, exclusion criteria such as calls from a public place and the high pass back rate means the number of ambulance journeys saved is only one quarter of that expected. Using Omega codes reduces the number of pass backs but these codes comprise only 5% of 999 workload and the other operational limitations means that potential ambulance journeys saved are reduced to less than 2% of total 999 call volume. Previous attempts to estimate the impact of transfer of 999 calls for telephone advice, including our own when this study was designed, have been based on the assumption that this would result in a high proportion of calls being referred away from the ambulance service system and in to primary care or self care. One of the most useful findings of this study is that it has shown that this is not the case and a high proportion of calls with low priority dispatch codes still require some form of immediate response or face to face medical assessment. The requirement for suitable referral processes to manage this process has already been discussed in detail. However there are two other considerations that need to be taken into account;

1. That these callers to the 999 service genuinely have a medical problem that requires emergency an ambulance response and therefore transfer for further assessment produces no advantages, or

2. Callers have a problem that does not require an emergency ambulance response but does require transport or a face to face contact with a health professional that cannot be arranged outside of the ambulance service within a suitable timeframe and hence an ambulance response becomes the default option.

In the first case we have demonstrated that, within the limitations of the current EMD and CAS systems some inappropriate referrals are inevitable. One option would be to abandon telephone referral until such time that a more sensitive system of triage and referral is developed. However, we have shown that where the system works well it is very effective, users are satisfied and even a small reduction in ambulance transports is cost-effective. Therefore even in its present state there are some

advantages. Furthermore continued call transfer can increase the empirical evidence base needed on which to base decisions about those calls that are suitable for transfer and those that are not. Increasing policy emphasis on telephone assessment

at the time of an emergency call means continued research in this area is now more essential.

In the second case there are a number of options that could improve the effectiveness of the service:

• A reappraisal of the role of nurse assessment of the management of 999 calls that moves away from the expectation that this will result in a cancelled ambulance journey for the majority of calls that receive this service and towards an enhanced level of triage that provides a more detailed

assessment of selected calls and hence a more considered response in terms of the use of ambulance service resources. This is a particularly pertinent issue for ambulance services. As the range of response options available increases so does the requirement for a more sensitive means of appraising the nature of calls and the most appropriate response. During the life of this study the development of the role of Emergency Care Practitioners (ECP’s) has moved forward at a rapid rate30. This role involves providing health care professionals with additional diagnostic, assessment and treatment skills that will allow them to provide an enhanced level of clinic care to patients in their home or to provide initial treatment and then refer patient to appropriate continuing care. Provision of this level of care at home in partnership with other services and hence reducing the number of patients taken

unnecessarily to hospital is another objective of “Taking Healthcare to the patient”. To date the majority of ECP’s are ambulance service professionals and have been employed by ambulance services. However there remains a difficulty for ambulance services in identifying calls suitable for an ECP response in much the same way that there is difficulty in identifying calls for a telephone advice response. The benefits of additional clinical assessment of calls may be as much concerned with aiding decisions about the deployment of an appropriate ambulance service response, including the effective use of ECP’s, as saving ambulance service journeys.

• The adoption of enhanced assessment to aid the decision making process around deployment of ambulance service resources raises issues about who and where this process takes place. In our study we have only investigated the use of nurses to provide further assessment and advice. However, there are questions around who should provide this assessment and it has been suggested that ambulance service personnel could fulfil this role. Earlier

further assessment11 and within the qualitative study we have conducted there is a view amongst ambulance service control room staff that, with additional training, they could take on this responsibility. The recognition transfer of calls for further assessment will still result in an ambulance response has lead to a reappraisal of where this assessment should take place. The early exit of the site in this study who had employed nurse advice within the ambulance service control room has made it difficult to compare this system with transfer to an NHS Direct service. There were however some clear issues identified. One is that the difficulties encountered within site 1 were more concerned with issues around management of staff rather than the call transfer processes. In this site there seemed to be fewer technical

difficulties in the call transfer process than when routed to a remote site. In the latter case there was sometimes confusion as to when a nurse was available, particularly if NHS Direct was busy and the ring fenced nurse was temporarily moved back to other call taking duties. In both cases the small number of calls transferred led to long periods when the assigned nurses were not utilised and therefore resources were wasted. Site 2 in this study changed the method of answering category C ambulance calls. Initially they used a ring fenced nurse but after moving premises and implementing an enhanced telephone system this allowed category C calls to be flagged up within the NHS Direct call queue and answered by the first available nurse thus dispensing with the need to provide a dedicated nurse for these calls. • If call taking roles are to be used effectively there seems to be some case for

considering expanded and more generic call taking roles. The additional workload for NHS Direct created by changes in primary care out of hours provision has resulted and a shift in priorities and management of ambulance service 999 calls is viewed by some to be a minor consideration within the current system. The combination of shift in priorities for NHS Direct and need to more appropriately deploy ambulance resources where cases cannot be referred elsewhere is increasingly leading ambulance services to consider keeping enhanced assessment and telephone advice within the ambulance service setting.

• The high pass back rate to ambulance services highlights problems within emergency care systems and in particular the referral of patients to

alternative services in within an acceptable timeframe. The true potential of referring some 999 calls for further assessment will only be achieved when

the appropriate service is available at the time of need. The development of properly integrated emergency care networks would seem to be a key issue in achieving this potential. Until then many calls will continue to be referred back to the ambulance service for non clinical reasons as illustrated in this study. Attempts are under way to better integrate services, for example an enhanced priority dispatch system which integrates with a clinical assessment system allowing better flow of information between systems with a resulting reduction in the amount of questioning needed of callers is already being utilised in the UK. However, the current system of ambulance service, NHS Direct and Out of Hours call handling systems for the most part remain separate but with referrals to each other. Whilst this system remains there remains the likelihood that some calls, as experienced by respondents to our questionnaire, will be successively transferred through a series of services only to end up back where they started. A properly integrated single

emergency care system call handling service would appear to be the logical way forward5. A recent report by the National Audit Office (NAO) supports this view and recommended that “emergency care networks should achieve maximum flexibility in the range of providers to which ambulance services transport or refer patients”31.

One further consideration is the expectations of the public in terms of what the 999 service will provide. The findings of our study have shown that, although some callers to the 999 service are aware that their call is not an emergency and are happy to receive an alternative service there are others who believe that if they call 999 and request an ambulance one should be sent without delay. The NAO report found a similar view expresses in the survey they conducted amongst members of the general public with only 30% of respondents agreeing that being connected to another service such as NHS Direct following a 999 call was acceptable31. The qualitative study found a degree of frustration amongst emergency medical

dispatchers in handling calls which they did not consider to be emergencies and the alternative service was seen by them to be a real step forward. Thus there are differences in the perceptions of emergency medical personnel and the public about what constitutes an emergency medical condition. This is likely to increase. One of the main sources of dissatisfaction in our study were delays in receiving an

ambulance and we have given some consideration to the clinical risk that may result from this response. However, one of the consequences of removing category C calls from response time performance standards is that some ambulance services have

implemented a policy of providing a response within one hour for these calls on the basis of the EMD code assigned. The system we have evaluated added an additional level of assessment which would result in an immediate ambulance response for some of these calls and therefore provided a “safety net” for ensuring delayed ambulance response was an appropriate action. Where a response time standard of one hour is employed without this additional assessment then clinical risk is likely to be increased and the consideration of safety becomes more important. It would be interesting to measure user satisfaction with this service change as one of the biggest challenges would appear to be changing public perception and expectations of the service they are likely to receive when they call 999.